Ep 415: Balancing bonus generosity using generative AI with Optimove’s Shai Frank and Dana Rausch

Joe Streeter (00:03.166)
Good afternoon and welcome to the iGaming Daily brought to you by OptiMove, the number one CRM marketing solution for the iGaming market. And today we have two guests from OptiMove. They've been here before. You know them both very well. Yeah, I will let them introduce themselves or reintroduce themselves to you though. Firstly, Shai, how are you? How is everything?

Shai Frank (00:24.578)
Hi, Joe. Thank you. Great being here. My name is Shai Frank. I'm Optimova Senior VP of Product. Happy to join you again.

Joe Streeter (00:33.002)
Brilliant and

Joe Streeter (00:37.594)
have you not been on Dayna? My fault, sorry. I'll debut.

Dana (00:39.166)
I have not. It's okay.

Shai Frank (00:43.724)
Yeah, it is.

Joe Streeter (00:47.56)
My fault, sorry.

Joe Streeter (00:52.532)
Good afternoon and welcome to the iGaming Daily brought to you by OptiMove, the number one CRM marketing solution for the iGaming market. And today we have two guests from OptiMove who are going to provide us some insight into CRM and in to finding the right balance when it comes to bonus generosity within iGaming. I will let them introduce themselves to you or reintroduce themselves in your case, Shy. We'll start with you. How is everything? How are you doing?

Shai Frank (01:21.336)
Hi, Joe. Thank you. Thank you for hosting me again. Great being back here. My name is Shai Frank. I'm OptiMove's Senior VP of Product. Great being you. Yeah, sorry.

Joe Streeter (01:31.786)
Brilliant.

What? Go, just go again. Say the intro again.

Shai Frank (01:37.452)
Yeah, my name is Chai Frank. I'm Optimoves SVP of products and great being here again.

Joe Streeter (01:43.658)
Brilliant, thanks Cheyenne. Dana, iGamer Daily debut for yourself. It's great to have you on the show. How is everything? Tell us more about your role at Optimoo.

Dana (01:47.86)
Yeah.

Dana (01:53.34)
Sure, I'm happy to be here. Thank you for welcoming me on for the first time. So I am Dana Rauch, strategic services team lead at OptiMove. So Shai is the man behind the product and I am the person behind all the custom solutions. So excited to chat a bonus thing today.

Joe Streeter (02:13.706)
Brilliant, so we've.

Dana (02:22.538)
Probably. I live on a busy street. No!

Dana (02:32.424)
They are all closed.

Dana (02:39.368)
Now of course.

Joe Streeter (02:39.818)
I'll give you that. And Dana and iGaming, Dana and iGaming daily debut for yourself. Thanks for joining us. If you could tell us more about your role and yeah, what you do at Optimo.

Dana (02:52.682)
Of course, excited to be here. Thank you for having me. So my name is Dana. As you said, I'm a strategic services team lead here at OptiMove. So quick explanation of my role. Shai is the guy behind all of our out of the box, the product build, and I deal more on the custom solution side of things. excited to talk to you today about bonusing.

Joe Streeter (03:15.786)
Brilliant. So we've kind of got the full machine here in terms of bonuses, so we can get some really good insights. And to kick us off, Shai, if you're able to, could you just remind the listeners of what overgenerosity is in marketing?

Shai Frank (03:31.64)
Sure, yeah. we talked in the previous episode at length about a lot of gaming operators being very generous with bonuses and promotions around player acquisition, around player retention. And what we've seen working with so many iGaming operators is that many times they are being very generous with promotions and bonuses and offers, even in cases where players might engage.

with a lower spend. So there are definitely opportunities to protect profit margins while still maintaining player experience and player engagement. And this is the topic that we want to elaborate on how to optimize bonus spend in a clever way, in a way that improves customer experience but also protects and improves profit margins.

Joe Streeter (04:28.018)
Where do you think that comes from, kind of that desire to be over generous with bonuses, just to get that sheer race to market share?

Shai Frank (04:38.55)
It probably stems from a combination of a race to market share and the genuine difficulty in running customer experience and engagement at scale and optimizing at scale. It involves a lot of segmentation and a lot of A-B testing and experimentation and analysis of results. And these are things that are generally hard to do. And when you want to operate at scale,

with large volumes of players, with large numbers of sports and casino and matches and different offers, sometimes it's just hard to scale, so it's easier to just throw out a very generous bonus to a very large audience of eligible players and get the results that you aspire to get.

While there's obviously room for optimization and improvements, it's not always easy. So it's probably one of the barriers to that.

Joe Streeter (05:40.744)
Yeah that makes sense and I guess over generosity is an issue felt across sectors but Dana it's a real challenge for our gaming operators right?

Dana (05:50.088)
Yeah, absolutely. I mean, we look at walking down the streets of New York where Shai and I are based. You see billboards for DraftKings, for Caesars giving away just loads of free money. So how are operators supposed to compete with that when we're, as you said, fighting for market share, fighting for that acquisition? That bonus is a real value to players, and players are definitely playing the field, looking for that best bonus.

deciding where they're going to play.

Joe Streeter (06:22.824)
And I guess it's a real issue too when it comes to protecting brand growth and ensuring brands can grow and continue to scale up like Shai said.

Dana (06:35.336)
Yeah, absolutely. We work with a lot of operators of a lot of different sizes. And we definitely see over generosity hurts operators of all sizes, I'm sure. Shai, you've seen this as well. We've had a lot of conversations with our clients on this. But it really hurts the small guys, right? It's difficult to keep up with the FanDuelz and the Caesars and the VET MGMs of the world.

Joe Streeter (07:02.056)
Yeah, I guess that's where it really pays to be creative with bonuses and with other incentives as well and not just kind of throw things at the wall as you say.

Shai Frank (07:14.296)
Yeah, absolutely. And just to follow up on what Dana just mentioned, we are talking with operators, some of them share with us some numbers about how long it takes to recover the acquisition cost of players, right? We are hearing from some of our clients in some of the states and some of the cohorts, incredible numbers, like it takes a decade to recover the acquisition cost. Or in some weird cases, think one of our clients told us,

a specific cohort in a specific state, they analyzed their specific month would take 68 years to recover. That's obviously an anecdotal number. It's not what they'd see across the board, but just I found it to be a very intriguing number. It just proves the point about the rise of acquisition costs in general, but also

the true price of acquiring players with very generous offers, but then are you losing money on these players and for how long? Is it okay to lose money on them for the first 12 months? Or if it takes you a decade to recover that cost, it's a losing business, right? So either you want to change your strategy or pull out. And we've seen some big operators pull out from some offers, from some...

Joe Streeter (08:28.586)
Yeah.

Shai Frank (08:39.496)
types of games from some states. because of that and if we can help and we believe that we can help with at least some of that aspect of optimizing spend, especially with player retention, then it's a profit.

Joe Streeter (08:58.314)
Okay, I just want to be, okay, go on, Diana, sorry.

Dana (08:58.986)
I just wanted to highlight something that Shai said that I feel like is very important that's easy to, I guess, not assign importance to, easy to miss. So he mentioned there's a specific cohort in a specific area, in a specific product type that has this very high acquisition ROI. What's important about this is that

every cohort, every state is very different. And this means that every bonus that we give out, we need to be considering the player level, right? What bonus is best for each player? And that's really the way to break the over-generosity habit, right, is by using that segmentation, by understanding your player base and really taking into account all of the cohort differences and making sure that you

really understand your player base to avoid over-generosity so that you're only bonusing where it really counts.

Joe Streeter (10:03.102)
That's absolutely fascinating and I guess it's kind of fair to say that the latest evolution of player is far smarter with bonuses than they were years ago. Is that a fair statement?

Dana (10:17.674)
Absolutely. I think this even spans industries, right? We have trained our players, we've trained our customers to expect a bonus. So this means that they are more likely to go elsewhere when there's not a bonus available. They're much sharper when it comes to looking for those offers because they know that they're out there. We've trained them to expect a certain type of offer.

Joe Streeter (10:47.604)
Yeah, that's fascinating to see how they've kind of evolved and why operators need to be far smarter with bonuses now. one thing I kind of want to clear up before I ask this next question is the difference between kind of bonus abusers and bonus seekers. One of you is able to kind of just define the difference between the two.

Shai Frank (11:12.558)
Dana?

Dana (11:13.898)
Sure. Yeah, a huge difference, honestly. And this kind of goes back to what I was saying, like really ensuring that you understand your player base, because these two players are very different, but they're often mistaken for one another, right? So a bonus abuser is someone who is coming to play with the intent of only playing when they receive a bonus, taking advantage of that bonus as much as they can.

and ultimately driving negative revenue over and over. A bonus seeker is someone that can actually have brand loyalty and they can actually come and play. still depositing money, but we consider them to be smart players. They're looking for the best offer. They want their dollar to stretch as far as possible where a bonus abuser, sure, they want their dollar to stretch as far as possible, but ideally they'll never put a dollar in.

So that's kind of the difference between the two of them. And again, why it's so important to really understand the nuances of your player base so that you're not accidentally punishing your bonus seekers as you would your bonus abusers.

Joe Streeter (12:26.344)
Yeah, well that was kind of my next question really. Like I guess it's so important to understand which player you're accommodating for, whether it's bonus abusers who you don't want on site, you need to detect bonus seekers or just players that are kind of, you know, tapping into a bonus. I guess, yeah, really it's vital that you understand which player is coming on board before you kind of get them, yeah, before the onboarding process takes place.

Dana (12:53.684)
Yeah, absolutely. Try. No, no, please.

Shai Frank (12:54.316)
Yeah, go ahead, Dino. All right. OK. I think the overarching notion here is the importance of segmentation and generating insights about your players. And the prerequisite of that is to tap into the player data, both historical for predictive model generation and also in real time for engagement. And then

how do you get access as a marketer, as an operator, how do you get fast access to the data and to the insights? So when you come up with a campaign or you come up with an offer, that you can be independent about it and you can do the right thing. So you can get the insights about the ratios of different players between bonuses and bets and deposits and their patterns and their gameplay history. And to see...

kind of to adapt the offer to the right player segment. It's really hard to do when you have data in silos or you are blind, or if you are relying on IT or BI teams to produce some reports to you because it really slows you down. And when business wants to move fast, we've seen operators time and again say, well, I cannot afford waiting two days now until IT or BI comes up with a list of players.

I need to get offer out the door. So then they'll make those sacrifices that Dana mentioned, right? They'll accidentally bundle the treatments for bonus seekers with bonus abusers. That would otherwise, if they had better access to data and faster insight to execution cycles, they would be able to differentiate between those audiences.

Joe Streeter (14:49.086)
Yes, Shai, I wanted to ask you more about that predictive modelling actually. That can be a real crucial tool, maximising that tool as well. mean, you mentioned they're breaking down the silos and ensuring that everybody is kind of collaboratively towards the same effort. That's crucial.

Shai Frank (15:06.914)
Yes, absolutely. And there are different models that can be applied in different stages of the personalization journey, if you will. Dana and her teams have been doing a lot of work on creating predictive models for classification of players, So classification of at-risk players for responsible gaming or early VIP detections or bonus abusers models and things along these lines.

These are models that are usually used when you do the segmentation. When you start your journey towards personalization, you start with segmentation. And those predictive models come into play there. You want to only get the players who match certain criteria. And you are using those models scoring to narrow down your audience. But there are a lot of machine learning and AI models further down the line to optimize your offers.

All right, so how do I, once I narrow down the segment that I want to interact with, there's still a lot of work that can be done to optimize the offer, to optimize the bonus, to optimize the reward type, the reward value based on things that are not necessarily player attributes, but rather their engagement in real time and the response to different offers. And these are things that some operators are trying to do manually with, you know, tests.

and experiments and A-B tests, but these are hard to scale. And technology these days allows us to do this on autopilots in an optimization way, right? And say, yeah, I'm throw into this segment that I narrowed down already three different offers, three different levels of generosity and let machine learning and artificial intelligence continuously optimize and...

based on the actual reaction, right? So Joe reacts really well to the medium level offer. So why offering more, right? Let's protect our margin and Joe is engaged and happy with our kind of medium tier offer. And that's great. And Shai needs more. And then a month later, NFL season ends and all of a sudden the medium offer is not enough to Joe anymore.

Shai Frank (17:31.566)
Because Joe is an NFL fan and he plays with little incentives and when NFL season ends, all of a sudden you need more incentive or else you're gone. And it's really hard to pick up on those signals if you do it manually. And if you have a system that continuously optimizes, it's just an autopilot. It caters for seasonality, for changes in player preferences, campaign fatigue, it's a real thing, right?

you can put things and let the system optimize for you. And this is one of the things that OptiMove can help operators do. And we've seen great success. And we have case studies with operators that have tried things like that. And we're able to save millions and millions of generosity dollars just by optimizing different offers automatically like that.

Joe Streeter (18:26.824)
Yeah, a lot to dive into there. you know, we've spoken quite a bit, but I feel like we could speak about all this for so much longer. may have to run into bonus time of the podcast, pardon the pun, but yeah, the, the, role of AI really can't be understated here because those margins are so important for operators, especially the smaller operators.

Dana (18:51.57)
Yeah, absolutely. I think something that Shy hit on, right, Joe, you're happy with your medium level bonus X times throughout the year, but you're a huge NFL fan, meaning you need a higher bonus during that year. This is where predictive modeling is crucial, right? Because it really allows us to really allows us to move fast.

Basically, right, because we're not waiting on a BI team, like Shai said as well, a BI team and IT team to dig up the insights for us every time we want to deliver a bonus. But instead we have this model that is constantly learning and training based on in the moment player data. And it's understanding how your player base is changing and telling you, hey, Joe was happy with a medium tier bonus last week, but this week he actually needs something else.

And it allows us to do that segmentation in a way that allows us to move at the speed of the player rather than at the speed of whatever team needs to get us the data. So that is a huge part of why AI and this predictive modeling is so important to the generosity issue.

Joe Streeter (20:09.172)
And that is the Optigini, right? I love that name by the way, the Optigini. I think it's a brilliant name, fantastic. That's the Optigini, right? That's what you guys are bringing to the table here.

Shai Frank (20:19.384)
Yeah, OptiGenie is an all-encompassing name for all of the AI applications inside of OptiMove, from predictive models to self-optimizing campaigns, self-optimizing journeys, generative AI, insight generation. It's all kind of under this umbrella of OptiGenie. And yes, we've been investing for years in these capabilities. It's a major part of our

product roadmap and strategy going forward. And with the proliferation of generative AI applications in the last two years, there's a lot of stuff that can be done from the obvious, you know, copy assistant and you want to run a compelling message, but also to less obvious applications like helping you get answers to some analytical questions that again,

help you reduce your dependency in analysts and BI teams. It's all this notion of we want marketing teams, we want operators to move fast in the pace of their players. And if you wanna see how did my campaigns that offer a specific type of a matchup bonus in the past three months worked for this player segment? This is an interesting BI question that's very important to answer when I'm planning my next weekend.

offer, it may take a long time to produce something like that, right? You need to have some analysts go and do some SQL queries in some databases and pull up some bonuses from bonus information from different systems and campaign engagement system information from different systems. These are hard questions to answer and may take a few days in normal circumstances. With generative AI these days, this becomes very easy. You just ask a question.

you ask OptiGene in plain language the questions that I asked, you wait five seconds and you get an answer. And it helps marketers, it helps CRM marketers that are usually not analysts, somehow bridge the skill gap of getting quick insights that would otherwise require an army of analysts to work for a few days and you get it within seconds and minutes so they can now use that insight.

Shai Frank (22:45.186)
to inform their decision and optimize their spend and narrow down the right segment and produce the right offer to the right segment in a way that improves player retention, improves customer experience, while also protecting profit margins. this is kind of, every time we, when we think from a product strategy perspective, this is kind of the North Star, right? So yes, increase player engagement, it's pretty important.

but also protect margins and improve margins. And how can we find the right balance between those two? So it's a win-win situation to operators and players alike. And luckily for us, advancements in technology in the last couple of years really unlocked a lot of these opportunities and we're trying as much as we can to tap into that.

Joe Streeter (23:33.406)
Yeah, sir. Incredibly exciting and also just a little bit scary just how developed the AI is. That's quite phenomenal really that you get such feedback from it and it's only going to grow as well. And I guess a couple more questions. Just, you spoke about the States, but as we see kind of new markets open up in 2025, fingers crossed it's looking likely or all roads are leading to various new markets.

Dana (23:39.306)
you

Joe Streeter (24:01.406)
How important is this going to be as operators kind of look to get that market share in these new pools of players?

Dana (24:12.426)
Crucial, I believe. We've talked about the high cost of acquisition, and maybe this is something that's necessary to gain market share. But where we can really improve this habit and cut back, protect our margins, is in retaining our customers, retaining our players. So we know that we'll always be competing for market share. We'll always be competing.

for that first deposit. But really, once we start learning about our players, once the market matures as well, we start to level out that competition will still be there. So we need to focus on retaining our players in a way that protects our margins, but also, you know, it's giving them the experience that they're looking for.

Shai Frank (25:06.166)
Yeah, and to follow up on that, we've seen in the past situations where operators are acquiring and reacquiring the same players again and again. And just because they spend so much energy in this land grab of acquiring players for the first time and not spending enough energy in smart strategies to retain them. So they turn away and I don't have to reacquire them.

And it's such a huge cost. And we've seen even, you know, even trivial stuff like election year, so Ed spent, you know, it's just becoming very expensive to put some ads in front of people because election year made Facebook and Google ads extremely expensive. So imagine that you have to like spend the same money again and again on the same player.

Joe Streeter (25:36.682)
Mm.

Shai Frank (26:02.646)
in the same state or even same player in different states. So the ability to stitch player identities and retain them over time is really important. And at OptiMove for years, our focus was always about retention and smart strategies for CRM and how can you increase lifetime value of players and minimize churn and improve customer experience and all that.

eventually drives down acquisition costs in general. So you have to overspend again and again.

Joe Streeter (26:35.455)
Yeah.

Such an interesting, unlikely consequence of the election, by the way. Social media ads, I never thought of that, but yes, so obvious really when you think about it. And yeah, so important too, to kind of what you guys do in stopping that reacquiring process, because that's just so costly. I understand that and I get why that's so crucial for operators. But yeah, just a final one for you guys to finish off.

The ceiling for this is so high, right? Moving into 2025. What does the future hold?

Shai Frank (27:15.31)
Do you want to ask this again?

Joe Streeter (27:19.198)
Yeah, The ceiling for this tech is so high. What does the future hold, Shai, looking into 2025?

Shai Frank (27:28.25)
I think the patterns that we've seen are probably going to likely to continue. Acquisition costs are getting higher and higher all the time. Fierce competition from major players, from major operators. Opening up of new states maybe can accelerate that process. But we've also seen advances in technology that can unlock new opportunities.

So we're looking at things like, yes, we talked about optimizing segments and then we talked about optimizing offers, right? So let's put three levels of generosity and let the system pick and choose the right level for each player. Our focus in the next year in OptiMove is going to be also around kind of the last mile optimization. So when we look at the actual content and the actual reward, how can we optimize there, right? So we can say,

You know what, it's not just a match-up bonus with a flat fee. Maybe it could be a percentage of your average wager amount for each player. And maybe that percentage, which is personalized, can also be AI-driven. So it's not just bet 20, get 50 flat. It's bet 20, get 50 % of your average amount.

And beyond that, maybe it's not 50 % for everyone. Joe gets 50%, Dana gets 40%. So if you think about it, if you look at optimization across the entire value chain of personalization, from the selection of segments, through the selection of offers and channels, all the way to the last mile reward type, reward currency, maybe some players

Joe Streeter (29:22.186)
some parents really need to build money. I'm some parents may work well with technical equipment.

Shai Frank (29:23.554)
really need real money bonusing, some players may work well with secondary currency, like experience points and spins or whatever it might be. So like we see in other industries, like you know as a consumer of retail, some people prefer free shipping and some people prefer like a $5 coupon. Equivalently, there are different player preferences. It's hard to tell in advance what each player might

want. And this is why you use AI for optimization. You try out, you see how people respond, you see what works and what's not, and kind of reiterate on that. And guess what? It's not enough because these preferences change over time. So you have to continuously optimize all the time because you cannot just run a one-time model of promotion affinity and determine for now and

Ongoing Joe is always going to prefer this type of promotion. That's not that's not true, right? So all these affinity models are are correct for the time that they are being Executed but they may change over time. So you need continuous optimization So at OptiMove you're constantly looking at kind of what kind of optimization we can apply That is that's ongoing that helps the marketers

Joe Streeter (30:24.189)
Yeah.

Shai Frank (30:45.08)
kind of free up their time and energy to strategize and not to run experiments manually all the time and tap into advancements in technology, especially around AI and find the right applications for generosity optimization. So that's, the future is exciting for us on that front.

Joe Streeter (31:03.37)
It really is exciting and I'm incredibly excited to see what is next for OptiMove and what is next for this tool as well. I think it's so crucial moving into 2025. But Dana, Shai, I thank you so much for your time today. It's been a really enjoyable and insightful podcast and hopefully look forward to catching up again soon as well with the latest in how you guys have developed the sector and continued.

to evolve CRM.

Dana (31:34.644)
Absolutely, thank you so much.

Joe Streeter (31:36.362)
Thank you.

Shai Frank (31:37.346)
Thank you Joe, always.

Ep 415: Balancing bonus generosity using generative AI with Optimove’s Shai Frank and Dana Rausch
Broadcast by